Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 May 1.
Published in final edited form as: Leukemia. 2019 Dec 3;34(5):1383–1393. doi: 10.1038/s41375-019-0640-4

Light Chain Amyloidosis induced inflammatory changes in Cardiomyocytes and Adipose Derived Mesenchymal Stromal Cells

Torri L Jordan 1,2, Khansaa Maar 2, Keely R Redhage 2, Pinaki Misra 2, Luis M Blancas-Mejia 2, Christopher J Dick 2, Jonathan S Wall 3, Angela Williams 3, Allan B Dietz 7, Andre J van Wijnen 4,5,2, Yi Lin 6,7,*, Marina Ramirez-Alvarado 1,2,*
PMCID: PMC7196017  NIHMSID: NIHMS1559004  PMID: 31796914

Abstract

Light Chain (AL) amyloidosis is a progressive, degenerative disease characterized by the misfolding and amyloid deposition of immunoglobulin light chain (LC). The amyloid deposits lead to organ failure and death. Our laboratory is specifically interested in cardiac involvement of AL amyloidosis. We have previously shown that the fibrillar aggregates of LC proteins can be cytotoxic and arrest the growth of human RFP-AC16 cardiomyocytes in vitro. We showed that adipose-derived mesenchymal stromal cells (AMSC) can rescue the cardiomyocytes from the fibril-induced growth arrest through contact dependent mechanisms. In this study, we examined the transcriptome changes of human cardiomyocytes and AMSC in the presence of AL amyloid fibrils. The presence of fibrils causes a ‘priming’ immune response in AMSC associated with interferon associated genes. Exposure to AL fibrils induced changes in the pathways associated with immune response and extracellular matrix components in cardiomyocytes. We also observed upregulation of innate immune associated transcripts (chemokines, cytokines, and complement), suggesting that amyloid fibrils initiate an innate immune response on these cells, possibly due to phenotypic transformation. This study corroborates and expands our previous studies and identifies potential new immunologic mechanisms of action for fibril toxicity on human cardiomyocytes and AMSC rescue effect on cardiomyocytes.

Keywords: light chain amyloidosis, immunoglobulin, amyloid, mesenchymal stromal cell, transcriptome, cell therapy

Introduction:

Light Chain (AL) amyloidosis is a systemic protein misfolding disease characterized by the abnormal proliferation of monoclonal plasma cells secreting immunoglobulin light chains (LC). Misfolding of LC leads to the formation of amyloid fibrils which deposit in vital organs leading to organ failure and death. Full length LCs are composed of a variable (VL) and a constant domain (CL), the former being the site of somatic mutations responsible for promoting misfolding and aggregation in AL amyloidosis [1]. Over half of AL amyloidosis patients present with cardiac involvement, which has the worst prognosis amongst all AL amyloidosis patients [2, 3]. Current treatments for AL amyloidosis target malignant LC producing plasma cells using chemotherapy or autologous stem cell transplant [4, 5]. These approved treatments can be poorly tolerated [5]. There are no approved therapies targeting the organ damage caused by the amyloid fibrils deposited [6].

Cellular models of AL amyloidosis developed by our laboratory and others [7-13] have provided useful tools to understand the mechanisms of AL fibril-mediated cellular damage. Unfortunately, developing an animal model of systemic amyloidosis (besides AA amyloidosis) with the physiologic phenotype has been unsuccessful [14]. Our laboratory has previously reported the cytotoxic effects of both soluble and fibrillar AL LC species on RFP-AC16, an immortalized human cardiomyocyte cell line [15]. We demonstrated that soluble LC proteins caused cellular dysfunction and apoptosis, whereas the fibrillar species caused cellular growth arrest [13]. In our previous work, we used two lambda proteins, AL-T05 (IGLV 1-41)[16] and Wil (IGLV 6-57) [11, 17, 18], the kappa protein AL-09 and the kappa germline (control) κI O18/O8 (IGKV 1-33), all germline donor sequences overrepresented in AL amyloidosis. Wil fibrils had the most toxic effect, causing complete cell growth arrest [13].

More recently, we reported the addition of adipose-derived Mesenchymal Stromal Cells (AMSC) to rescue RFP-AC16 from growth arrest caused by Wil fibrils [12]. AMSC are multipotent cells isolated from adipose tissue of consenting, healthy donors. These cells have immunomodulatory, anti-inflammatory, and paracrine tissue remodeling properties [19]. The rescue effect seen with AMSC in our co-culture was contact dependent [12]. We also saw changes in the expression of extracellular matrix proteins and cell adhesion proteins in RFP-AC16 cells exposed to Wil fibrils.

Using our AMSC rescue model, we sought to characterize transcriptome changes in RFP-AC16 cardiomyocytes when cultured with Wil fibrils, AMSC (COCX), or co-cultured with Wil fibrils and AMSC (COCX Wil). We have shown the involvement of complement pathway activation as a response to amyloid fibrils for the first time in AL amyloidosis. In this study, we observed upregulation of cytokine and chemokine transcripts from both RFP-AC16 and AMSC in response to the presence of amyloid fibrils. This work highlights potential immune mechanisms by which amyloid fibrils cause significant phenotypic changes within the cardiomyocytes, induce their toxicity, and how AMSC may play a role in cardiomyocyte rescue.

Methods:

Protein Preparation

Wil immunoglobulin light chain variable domain (VL) λ6 protein and Wil fibrils were prepared and generated as previously described [11, 17, 18]. The fibrils were analyzed using the methods described in Misra et al [20].

Cell Culture

Mesenchymal stromal cells were derived from lipo-aspirates obtained from a consenting healthy donor (MSC 211) with approval from the Mayo Clinic Institutional Review Board, expanded in vitro, and authenticated following the protocol by Dudakovic et al. [21].

RFP-AC16 human primary ventricular cardiomyocytes were obtained from Dr. Mercy Davidson at Columbia University [15]. AC16 cells co-transfected with lentivirus expressing red fluorescent protein (RFP) in the nucleus was used (RFP-AC16). Cells were maintained as reported before [12].

RFP-AC16 cells are not listed in the database of commonly misidentified cell lines maintained by the International Cell Line Authentication Committee. Cell morphology was checked before each experiment, and the number of cell passages after thawing was limited to 20. Authentication of RFP-AC16 and cell line contamination testing is done every 6 months in our laboratory by Western blot and polymerase chain reaction.

AMSC and RFP-AC16 COCX Model

RFP-AC16 (12 500 cells/well) were seeded in a 24-well Corning polystyrene plate on day −1 and allowed to adhere overnight at 37°C (in Dulbecco’s Modified Eagle’s Medium/f12 Medium, Life Technologies, Carlsbad, CA). The following day (day 0) media was replaced with fresh media, and AMSC (2 500 cells/well), Wil fibrils (6 μM), or both were added and incubated 37°C, 5% CO2 for 150 hours with monitoring using the IncuCyte.

Isolation of AMSC and RFP-AC16 in the COCX model

Flow cytometry sorting was done on an Aria 4 Laser. Samples were sorted via expression of mKate2, Forward Scatter, and Side Scatter. Cells were sorted at 4°C, under sterile conditions with a 100 μm nozzle. Sorted cells were immediately suspended into lysis buffer supplied by the Qiagen RNeasy kit and RNA was extracted following sorting.

RNA Sequencing and Computational Analysis

RNA extraction was done according to the manufacturer’s instructions using the Qiagen RNeasy kit from RFP-AC16 cultured alone (300 000 cells), AMSC cultured alone (300 000 cells) or sorted cells from the COCX as described above.

RNA preparation for sequencing was done by the Genome Analysis Core at Mayo Clinic in Rochester, MN using Standard TruSeq v2 for mRNA (Illumina).

Samples were read on an Illumina HiSeq 4000 as paired end reads for 100 cycles. Sequence data was processed using MAP-RSeq (v1.2.1) and bioinformatics work flow done by the bioinformatics core at Mayo Clinic, where expression data was normalized using reads per kilobase per million (RPKM) on Human Genome 19. RPKM were calculated from the RNA sequencing data sets and analyzed to assess differential gene expression between conditions. Genes with RPKM >0.3 cutoff were used in the computational analyses. Differentially expressed genes were determined using ± 2 Log2 fold change (FC). Volcano plots were generated using GraphPad Prism v7.03 for Windows. Heat maps and hierarchical analysis were performed using Morpheus (Broad Institute). Gene Ontology (GO) Term networks and protein-protein interaction networks were assessed using STRING Database v10.5 [22]. Principal Component analysis was assessed using Cluster Vis [23].

Data Sharing Statement:

For original data contact ramirezalvarado.marina@mayo.edu

Results

Validation of Cell Isolation and RNA sequencing methods

We designed our experiments to identify transcriptome changes observed in RFP-AC16 in the presence of Wil fibrils and in COCX Wil as previously reported (Supplemental Figure 1A) [12]. Flow parameters for sorting of AMSC and RFP-AC16 are shown in Supplemental Figure 1B. We observed that in unbiased hierarchical clustering and principal component analysis of all the cell conditions, the biggest clustering group among all the conditions was cell type: AMSC clustered away from RFP-AC16 of all experiment conditions except for AMSC COCX (Supplemental Figure 2A,B). We observed that AMSC expressed classical surface antigens CD105, CD73, and CD90 (ENG, NT5E, THY1) in high RPKM, and did not express (or expressed in very low RPKM) CD14, CD34, and CD45 (PTPRC) as reported previously [24].

Next, we wanted to determine if there are changes in the transcriptome that occur as a result of passing cells through a sorter. The RFP-AC16 alone and flow-sorted RFP-AC16 clustered closely to each other, suggesting that any gene expression changes induced by the flow sorting is negligible relative to the experimental conditions studied (Supplemental Figure 2A,B). Indeed, a comparison between Sorted RFP-AC16 and Unsorted RFP-AC16 showed that out of 12 110 genes (with an Average RPKM of > 0.3), 0.18% were up or down regulated by greater than 1 Log2 FC. Of those 22 genes, only 4 were greater than ±2 Log2 FC. We determined that RFP-AC16 do not drastically change their transcriptome when run through a FACS Sorter and we elected to only compare sorted RFP-AC16 transcripts FC with the other experimental conditions. We also investigated transcriptome changes between sorted and unsorted AMSC. We observed that out of 11 800 genes (with an average RPKM>0.3), 3.0% were up or down regulated by greater than 1 Log2 FC. The majority of the genes were upregulated, with only 20 being greater than 2 Log2 FC. The genes that were upregulated here did not match with any genes identified in our subsequent analysis and therefore not likely to confound the results. We concluded AMSCs are more likely to be affected by FACS sorting, and as in the case of the data from RFP-AC16, we only used sorted AMSC for comparison with sorted AMSC experimental conditions.

However, when we examined cells isolated from COCX conditions, RFP-AC16 isolated from COCX still clustered closer to other RFP-AC16 conditions, but the AMSC sorted from COCX conditions clustered closer to RFP-AC16 (Supplemental Figure 2A, B). This was not surprising, given that we intentionally set up the COCX condition with small AMSC to RFP-AC16 ratio. Thus it is highly likely that flow sorting was sufficient to enrich for a more pure population of RFP-AC16 cells but inadequate to enrich for a pure AMSC population. We also note that AMSC have auto-fluorescence on the fluorescent channel (mKate2) used to sort cells that have an overlap with the lower range of dimly fluorescent red RFP-AC16 cells, causing incomplete sorting of AMSC from RFP-AC16 cells (Supplemental Figure 1B). Indeed, when we examined for genes that were highly expressed in AMSC and not expressed in RFP-AC16, we found that these genes were not expressed in flow sorted RFP-AC16 COCX and were also not expressed in AMSC sorted from COCX (Supplemental Figure 2C). In contrast, genes highly expressed in RFP-AC16 and not expressed in AMSC were found in AMSC isolated from COCX with AC16 and with comparable expression level to RFP-AC16 (Supplemental Figure 2C). Therefore, for the analysis of this paper, we excluded AMSC sorted from COCX conditions but included RFP-AC16 isolated from COCX conditions. With this information, we are confident that the cells in our study, the cell culture conditions, and the sorting methods that we chose to carry out the RNA sequencing and subsequent experimental analysis will offer us accurate results about the changes in the transcriptome.

Transcriptome Changes of AMSC in the presence of Wil Fibrils

We showed previously that AMSC do not change their morphology and do not exhibit growth arrest when incubated with Wil fibrils, but did not assess any possible changes in the transcriptome that will not be reflected in overall morphology or growth changes [12]. Our transcriptome comparison between AMSC cultured with Wil fibrils and without Wil fibrils showed 8.5% of genes have greater than 1 Log2 FC (843 genes upregulated, 276 downregulated). Of the upregulated genes, 105 had a Log2 FC greater than 2. STRING networks noted a handful of genes connected in a related protein network (Figure 1A). From these networks, of note are the following genes: IL8, CD36, CCL5, IL1B, and CXCL10. These genes are all implicated in cytokine, immune, and defense related responses and pathways (Figure 1B). Other notable genes are SQSTM1, SOD2, GPNMB, SCUBE2, and ICAM1. The majority of genes encode membrane bound surface proteins playing roles in cellular adhesion and response to stimulus, while SOD2 is a gene encoding for a superoxide dismutase, associated with oxidative stress response. According to GO Terms, upregulated genes included response to external stimulus, negative regulation of cellular proliferation, regulation of cell migration, granulocyte chemotaxis, and leukocyte chemotaxis (Figure 1C). The 25 downregulated genes with less than −2 Log2 FC showed connections implicated in FOS/Jun signaling pathway (Figure 1D). Genes of note are FOS, EGR1, DUSP1, JUNB, FOSB, and IER2 (Figure 1E). The downregulated GO Terms included response to organic cyclic compound, cellular response to organic substance, and positive regulation of gene expression (Figure 1F). Collectively, our results show that AMSC undergo transcriptome changes with upregulation of immune related genes, as well as proteins on the surface of the AMSC that could potentially play a role in the AMSC ability to rescue RFP- AC16 from amyloid growth arrest.

Figure 1. AMSC cultured with Wil fibrils shows an immune based priming response.

Figure 1.

Open in a new tab

(A) Protein-Protein interaction network of upregulated genes shown have greater than 2 Log2 FC in AMSC Wil cultures. STRING Analysis was performed using high confidence (0.700), and took into account data from curated databases (light blue), experimentally determined (pink), gene co-expression (black), and gene co-occurrence (dark blue). Nodes not connected were not shown. (B) Immune and surface marker related proteins are upregulated when AMSC are incubated with Wil fibrils. Average RPKM Values of select upregulated genes in AMSC Wil cultures compared to unsorted AMSC. (C) GO Terms of AMSC Wil incubation gene upregulations focus on immunology and response to stimulus. GO Terms based on the STRING Analysis. The analysis was performed with high confidence (0.700), and at significant false discovery rate. (D) Protein-Protein interaction network of downregulated genes shown have greater than 2 Log2 FC in AMSC Wil cultures. STRING Analysis was performed high confidence (0.700), and took into account data from curated databases (light blue), experimentally determined (pink), gene co-expression (black), and gene co-occurrence (dark blue), and text mining (yellow). (E) Average RPKM Values of select upregulated genes in RFP-AC16 Wil cultures compared to unsorted RFP-AC16. (F) GO Terms of downregulated genes from AMSC Wil experiments focus on response to compounds and stimulus. GO Terms based on the STRING Analysis. The analysis was performed with high confidence (0.700), and at significant false discovery rate.

Transcriptome changes in RFP-AC16

PCA analysis of all the experimental conditions for RFP-AC16 showed that RFP-AC16 COCX with AMSC clustered away from RFP-AC16 alone, those exposed to Wil, and RFP-AC16 exposed to Wil and COCX with AMSC clustered away from cells exposed to Wil alone or COCX with AMSC without Wil (Figure 2A). Distinct clustering of the different experimental conditions supports involvement of differential biologic process or differential signaling of the biologic process. Scree plot showed that the first three PCAs contributed to more than 80% of the variations for the clustering (Supplemental Figure 2D). Thus the genes with the highest PCA loading and more than ± 2 FC compared to RFP-AC16 alone were used for hierarchical clustering (Figure 2B). Among these genes, 98 (5%) genes had changes that were unique to incubation with Wil while COCX Wil had 202 (10.3%) unique gene changes. There were 454 (23.1%) of the genes with changed expressions that are shared between Wil fibrils and COCX Wil fibrils, within which 261 (13.3%) were shared among all three experimental conditions (Figure 2C).

Figure 2: Principal Component analysis of RFP-AC16 cultures shows differential clustering separation from COCX with AMSC with and without Wil fibrils.

Figure 2:

Open in a new tab

(A) PCA map of PCA1vsPCA2 and PCA2vsPCA3 for RFP-AC16 culture conditions are shown here. Colors of conditions are noted in the figure legend. (B) Hierarchical clustering of RFP-AC16 culture conditions is shown in this heat map. All genes that had up or down-regulation of 2 FC or more for at least one comparison between any experimental conditions and RFP-AC16 alone were included for hierarchical clustering analysis in ClusterVis. (C) Venn diagram of shared and unique genes with changes in RNA expression among the experimental RFP-AC16 conditions are shown here.

Transcriptome changes of human cardiomyocytes with Wil Fibrils

RNAseq analysis was performed using RPKM values of >0.3 for genes in at least one cell type (12 000 genes total). First, we calculated the Log2 FC of unsorted RFP-AC16 with Wil fibrils compared to unsorted RFP-AC16. 586 genes were upregulated more than 1 Log2 FC and 253 genes downregulated less than −1 Log 2 FC; in total, 7.0% of genes (Figure 3A).

Figure 3. RFP-AC16 cultured with Wil Fibrils shows upregulation of immune related genes.

Figure 3.

Open in a new tab

(A) Volcano plot of RFP-AC16 Wil vs unsorted RFP-AC16 RNA transcripts greater than 0.3 AVG RPKM of cells, compared using a Log2 FC. (B) Protein-Protein interaction network of upregulated genes shown have greater than 2 Log2 FC in RFP-AC16 Wil vs unsorted RFP-AC16. STRING Analysis was performed with high confidence (0.700), and took into account data from curated databases (light blue), experimentally determined (pink), gene co-expression (black), and gene co-occurrence (dark blue). Nodes not connected were not shown. (C) Selected transcripts upregulated in RFP-AC16 Wil compared to unsorted RFP-AC16. (D) GO Terms based on the STRING analysis of 2 Log2 FC upregulated genes between RFP-AC16 Wil and unsorted RFP-AC16. The analysis was performed at high confidence (0.700), and at significant false discovery rate.

We queried Gene Ontology (GO) Terms and protein-protein networks (Using STRING) with the top up and down regulated genes with greater than 2 Log 2 FC in RFP-AC16 Wil. There were no protein-protein interactions determined by STRING in down regulated genes, the 91 upregulated genes with greater than 2 Log 2 FC had interactions (Figure 3B). Of specific note are numerous immunological response related genes: CXCL2, IL1B, CXCL3, IL8, IL11, IL6, CSF3, CXCL1, C3, and CSF2 (Figure 3C). These genes are all associated with chemokine, cytokine, immune, and defense related responses. Other genes of note include, MT2A, G0S2, ANPEP, MMP1, and MMP3. MMP1 and MMP3, and ANPEP are matrix metallopeptidases and proteases associated with extracellular matrix break down or other proteolytic cleavage. MT2A plays a role as an anti-oxidant, and G0S2 is the gene responsible for the switch moving from G0 to G1 phase during cell cycle and promotes apoptosis through the BCL2 pathway. GO Terms focused around immune response, defense response, response to (foreign) organism, and response to cytokine (Figure 3D). Additionally, GO Terms around cell regulation of locomotion, cell migration, localization, and proliferation were noted, corroborating with our previous research. Overall, we observed a trend indicating that RFP-AC16 human cardiomyocytes express defense response genes against the amyloid fibrils.

Transcriptome Changes of COCX Model in response to Wil fibrils

To understand transcriptome changes occurring in RFP-AC16 and AMSC when in COCX Wil, we compared RFP-AC16 Wil with RFP-AC16 COCX Wil (Figure 4A). This rescue comparison showed about 3.4% of genes up- or down- regulated greater than 1 Log2 FC (350 upregulated, 108 downregulated). Comparing sorted RFP-AC16 Wil to RFP-AC16 COCX Wil, STRING analysis of protein networks showed interactions between the genes upregulated with greater than 1 Log2FC (Figure 4B). There was a common theme of upregulated genes between the RFP-AC16 Wil and RFP-AC16 COCX Wil, showing many of the same genes being upregulated (G0S2, IL1B, IL8, IL11, IL24, CXCL1, CSF3, CSF2 and, C3). Between the RFP-AC16 with Wil and the RFP-AC16 COCX Wil, we see an upregulation of certain genes, such as OAS1, OAS2, OAS3, IL33, MX1, and MX2 (Figure 4C). GO Terms of the upregulated genes between Wil and COCX Wil included immune response, regulation of metabolic processes, signal transduction, response to stress, and regulation of biological processes (Figure 4D). Downregulated transcripts between Wil and COCX Wil included WISP2, CCL11, CFI, and C3 (Figure 4C). STRING analysis shows the connectivity of these genes at −1 Log2 FC or less (Figure 4E). These genes were seen upregulated in the RFP-AC16 Wil COCX, but decreased when AMSCs were added to the culture. This differential expression from AMSCs suggests a rescue capability on the cardiomyocytes from the effect of Wil fibrils. Some of the GO terms associated with these downregulated transcripts include cell differentiation, regulation of cell communication, developmental processes, and response to stimulus (Figure 4F).

Figure 4. Addition of AMSC for AC16 COCX Wil show differential changes in immune and metabolic pathways.

Figure 4.

Open in a new tab

(A) Volcano plot of RFP-AC16 COCX Wil vs RFP-AC16 Wil transcripts greater than 0.3 AVG RPKM, compared using a Log2 FC. (B) Protein-Protein interaction network of 1 Log2 FC upregulated genes in AC16 COCX Wil cultures compared to AC16 Wil is shown here. STRING Analysis was done at high confidence (0.700), and took into account data from curated databases (light blue), experimentally determined (pink), gene co-expression (black), and gene co-occurrence (dark blue). Nodes not connected were not shown. (C) Log2 FC of selected up and down-regulated genes in RFP-AC16 COCX Wil condition compared to RFP-AC16 Wil. (D) GO Terms of upregulated genes in RFP-AC16 COCX Wil compared to RFP-AC16 Wil. The analysis was performed at high confidence (0.700), and at significant false discovery rate. (E) Protein-protein interaction network of downregulated genes shown have greater than 1 Log2 FC in RFP-AC16 COCX Wil compared to RFP-AC16 Wil. STRING Analysis was performed with high confidence (0.700), and took into account data from curated databases (light blue), experimentally determined (pink), gene co-expression (black), and gene co-occurrence (dark blue). (F) GO Terms of downregulated genes in RFP-AC16 COCX Wil compared to RFP-AC16 Wil. The analysis was performed at high confidence (0.700), and at significant false discovery rate.

Discussion

In this paper, we report transcriptomic data on the impact of Wil amyloid fibrils on human cardiomyocytes, AMSC, and the effect of AMSC on cardiomyocytes exposed to Wil fibrils. Interestingly, while different biologic response in terms of cell growth were seen when AMSC and RFP-AC16 were exposed to Wil alone, immune related pathways were affected in both cell types. These immune pathway changes suggest that RFP-AC16 cells are capable of participating or responding to immune responses.

Complement factor C5b9 upregulation has been observed in association with amyloid deposits and implicated in both neuronal and muscular damage of patients with systemic amyloidosis [25-27]. The complement pathway has also been linked to inflammatory responses to Aβ amyloid in Alzheimer’s disease, specifically complement protein C1q [28]. C3 upregulation in Alzheimer’s disease has previously been noted in microglial cells; the deficiency of C3 has some protective capability against neurodegeneration in murine models [29-31]. Given the upregulation of C3 in our in-vitro model and down-regulation with the addition of AMSC, we believe complement may play a role in the signaling cascade triggering cell damage caused by amyloid deposition in cardiac AL amyloidosis. C3 could also play a role initiating an inflammatory response both through the innate immunity pathway and through the inflammasome.

The upregulation of IL-1β in our COCX experiments is interesting because it points towards the potential for inflammasome activation, which has been observed in various amyloid diseases such as Alzheimer’s, Type 2 Diabetes, and amyotrophic lateral sclerosis (ALS) [32, 33]. In Alzheimer’s disease, fibrillar species of Aβ (not the soluble precursor or intermediates), lead to early and high expression of IL-1β, which indicates a separate mechanism of action by fibrillar species compared to soluble species, in agreement with our previous report [32]. Reactive oxygen species and lysosomal damage (indicated previously in AL amyloidosis research from our laboratory and others [11, 33, 34] have been implicated as a common trigger for the NLRP3 inflammasome. The activation of NLRP3 may be involved in the sensing of protein aggregates and the eventual immune response that follows. We propose this may be one of the main cellular damage responses in AL amyloidosis.

The RNA sequencing experiments elucidated numerous cytokines and chemokines upregulated in response to amyloid fibrils and the addition of AMSC in the COCX. Many of the cytokines and chemokines identified (CSF3, CSF2, CXCL1, CXCL3, CXCL6, IL8, CCL3) are implicated in innate response, specifically neutrophil or granulocyte responses. CXCL1, CXCL3, IL8 (also known as CXCL8) and CXCL6 are all members of the CXCL8/IL8 family, and are classically known for neutrophil activation and tissue damage [35]. When members of this chemokine family are present, neutrophil-dependent inflammation and response generally follow. CSF3 and CSF2 encode GCSF and GMCSF, which are extremely important in the life cycle of a neutrophil or other granulocyte/monocytes. GCSF and GMCSF are major factors in regulating neutrophil proliferation, inflammation, survival, and mobilization [36]. CCL3 is a pro-inflammatory cytokine playing a role in recruitment and activation of granulocytes, macrophages, and monocytes [37]. The increase in innate related transcripts sheds a light on the type of immune response occurring when amyloid is present, and perhaps a mechanism by which the immune system responds and subsequently helps or harms the cardiomyocytes.

The transcriptional shift of RFP-AC16 cardiomyocytes to a more immune-related state may signal a phenotypic shift in the cell. This has previously been suggested in a renal AL amyloidosis cell model using human renal mesangial cells [38, 39]. Our data reveals that RFP-AC16 cardiomyocytes expressing immune related transcripts, something typically expressed in endothelial cells. In reference to the phenotypic transition seen in renal mesangial cells, we propose our cardiomyocytes may be undergoing a phenotypic shift, after interacting with fibrils, towards an endothelial state which is more prone to acting in an immunologic fashion.

When AMSC are cultured with Wil fibrils, we note our GO terms show immune response, and we saw an upregulation in immune related and surface marker upregulation genes. It is suggested that AMSC hone to inflammation, where they are able to exert an immune effector or suppressive response in the region. It is also hypothesized they may have an anti-apoptotic effect [40]. The presence of interferon and response pathways suggest a potential for ‘priming’ AMSC in response to fibrils prior to utilization in our COCX rescue model. AMSC priming has been used in cancer and immunosuppressive contexts; we hope to further explore the priming of AMSC in our rescue model to elucidate the mechanisms of action by our AMSC in context of AL amyloidosis [41, 42]. Between the RFP-AC16 Wil and RFP-AC16 COCX Wil conditions we see upregulation of the OAS family of genes (1,2,3). These genes are associated with a viral based innate immune response [43]. We also see an upregulation of MX1 and MX2, which are proteins that also participate in antiviral responses, and type 1 and 2 interferon responses [44].

Although our analysis points out the upregulation of immune transcripts, this analysis does not make it clear if the response to the presence of Wil fibrils is beneficial or detrimental. Due to the lack of research on the immune response in AL amyloidosis, it is difficult to ascertain what the role of the upregulated transcripts are without further studies. Future research will address these questions related to the role of the immune response, defense related transcripts, and upregulation of the ECM and surface marker transcripts in AL pathophysiology. We also intend to compare upregulation of transcripts with the corresponding protein synthesis.

AL amyloidosis has typically been studied in-vitro due to the lack of a successful animal model, particularly an animal model with the cardiac manifestation phenotype that is seen in patients [14]. Biopsy cardiac tissues are difficult to obtain; therefore, cellular models of AL amyloidosis have an increased importance in being one of the few models for studying damage and toxicity caused by AL fibrils.

We have used RNAseq as a way to corroborate our previous results, but also to begin to elucidate new mechanisms by which amyloid can impact both cardiomyocytes and AMSC. We have verified our sorting methods, and note the small (0.18% in RFP-AC16, 3.0% in AMSC) up or down regulation of genes in cells after they have been sorted. This is important for future genomic base studies, to ensure that responses seen are not caused by the sorting but caused by the conditions that are being examined. We acknowledge that our analysis does not have a control fibril based on our previous analysis [12, 13]. Future studies from our laboratory will include the RNAseq analysis using control proteins.

Our contribution to the field is that our RNAseq study emphasizes the importance of an emerging area of research in amyloid diseases: understanding the etiology in the context of immunological systems. Our RNAseq implicate the need for an immunologic survey of AL amyloidosis, but other work has indicated an immunological aspect to other amyloidogenic diseases. The recent discovery that the Alzheimer’s Disease peptide Aβ can act an antimicrobial peptide (AMP) against bacteria, fungi, and viruses [45] could be interpreted as supporting evidence to an infection-associated hypothesis for Alzheimer’s disease. Several laboratories have shown that microbial infection increase the synthesis of this AMP [46]. These investigators propose that Aβ as an AMP will be beneficial on first microbial challenge but will become progressively detrimental as the infection becomes chronic and reactivates from time to time [46]. With the resurgence of the infection driven hypothesis for Alzheimer’s, it is important to understand tissue damage in amyloidosis beyond the misfolding processes.

In this paper we have begun to examine the changes in the transcriptome when immortalized adult human cardiomyocytes are COCX with Wil fibrils, AMSC, or a combination of the two. Through this, we were able to hypothesize that the immune system may play an important role in disease progression and response, and that the cardiomyocytes themselves are beginning the immune response process, potentially via a phenotypic shift. This immune response seems to be focused around complement activation and mobilization of a granulocyte response. Overall, our findings stress the importance of looking at cardiac AL amyloidosis through an immunologic lens.

Supplementary Material

supplementary material

Acknowledgements:

We thank the staff of the Medical Genome Facility Expression Core for carrying out the RNAseq analysis and the staff of the Flow Cytometry Core for their assistance. We also thank Michael Bergman, Shawna Cooper, Christopher Parks, and Christopher Paradise for their contributions to this project. TLJ is a graduate student at Mayo Clinic Graduate School of Biomedical Sciences. This work is submitted in partial fulfillment of the requirement for the PhD program.

This study was supported in part by NIH R01 GM 128253, the Mayo Foundation, and the generous support of amyloidosis patients and their families.

Abbreviations:

AL

Amyloidogenic Light Chain

AMSC

Adipose-derived Mesenchymal stromal Cell

COCX

Co-culture of RFP-AC16 cardiomyocytes with AMSC

COCX Wil

Co-culture of RFP-AC16 with AMSC and Wil fibrils

FC

fold change

LC

Immunoglobulin light chain

RFP

Red Fluorescent Protein

RPKM

reads per kilobase million

Footnotes

Conflict of Interest Disclosure:

The authors declare no competing financial interests.

References

  • 1.Baden EM, Sikkink LA, Ramirez-Alvarado M. Light chain amyloidosis-current findings and future prospects. Current Protein and Peptide Science. 2009;10(5):500–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Blancas-Mejía LM, Ramirez-Alvarado M. Systemic amyloidoses. Annual review of biochemistry. 2013;82:745–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Falk RH. Diagnosis and management of the cardiac amyloidoses. Circulation. 2005;112(13):2047–60. [DOI] [PubMed] [Google Scholar]
  • 4.Merlini G CyBorD: stellar response rates in AL amyloidosis. Blood. 2012;119(19):4343–5. [DOI] [PubMed] [Google Scholar]
  • 5.Mikhael JR, Schuster SR, Jimenez-Zepeda VH, Bello N, Spong J, Reeder CB, et al. Cyclophosphamide-bortezomib-dexamethasone (CyBorD) produces rapid and complete hematologic response in patients with AL amyloidosis. Blood. 2012;119(19):4391–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Gertz MA. Immunoglobulin light chain amyloidosis: 2016 update on diagnosis, prognosis, and treatment. American journal of hematology. 2016;91(9):947–56. [DOI] [PubMed] [Google Scholar]
  • 7.Trinkaus-Randall V, Walsh MT, Steeves S, Monis G, Connors LH, Skinner M. Cellular response of cardiac fibroblasts to amyloidogenic light chains. The American journal of pathology. 2005;166(1):197–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Brenner DA, Jain M, Pimentel DR, Wang B, Connors LH, Skinner M, et al. Human amyloidogenic light chains directly impair cardiomyocyte function through an increase in cellular oxidant stress. Circulation research. 2004;94(8):1008–10. [DOI] [PubMed] [Google Scholar]
  • 9.Sikkink L, Ramirez-Alvarado M. Cytotoxicity of amyloidogenic immunoglobulin light chains in cell culture. Cell death & disease. 2010;1(11):e98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Levinson RT, Olatoye OO, Randles EG, Howell KG, DiCostanzo AC, Ramirez-Alvarado M. Role of mutations in the cellular internalization of amyloidogenic light chains into cardiomyocytes. Scientific reports. 2013;3:1278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.McWilliams-Koeppen HP, Foster JS, Hackenbrack N, Ramirez-Alvarado M, Donohoe D, Williams A, et al. Light chain amyloid fibrils cause metabolic dysfunction in human cardiomyocytes. PLoS One. 2015;10(9):e0137716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Lin Y, Marin-Argany M, Dick CJ, Redhage KR, Blancas-Mejia LM, Bulur P, et al. Mesenchymal stromal cells protect human cardiomyocytes from amyloid fibril damage. Cytotherapy. 2017;19(12):1426–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Marin-Argany M, Lin Y, Misra P, Williams A, Wall JS, Howell KG, et al. Cell Damage in Light Chain Amyloidosis: FIBRIL INTERNALIZATION, TOXICITY AND CELL-MEDIATED SEEDING. Journal of Biological Chemistry. 2016;291(38):19813–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Buxbaum JN. Animal models of human amyloidoses: Are transgenic mice worth the time and trouble? FEBS letters. 2009;583(16):2663–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Davidson MM, Nesti C, Palenzuela L, Walker WF, Hernandez E, Protas L, et al. Novel cell lines derived from adult human ventricular cardiomyocytes. Journal of molecular and cellular cardiology. 2005;39(1):133–47. [DOI] [PubMed] [Google Scholar]
  • 16.Poshusta TL, Katoh N, Gertz MA, Dispenzieri A, Ramirez-Alvarado M. Thermal stability threshold for amyloid formation in light chain amyloidosis. International journal of molecular sciences. 2013;14(11):22604–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Wall J, Schell M, Murphy C, Hrncic R, Stevens FJ, Solomon A. Thermodynamic Instability of Human λ6 Light Chains: Correlation with Fibrillogenicity. Biochemistry. 1999;38(42):14101–8. [DOI] [PubMed] [Google Scholar]
  • 18.Wall JS, Gupta V, Wilkerson M, Schell M, Loris R, Adams P, et al. Structural basis of light chain amyloidogenicity: comparison of the thermodynamic properties, fibrillogenic potential and tertiary structural features of four Vλ6 proteins. Journal of Molecular Recognition. 2004;17(4):323–31. [DOI] [PubMed] [Google Scholar]
  • 19.Bernardo ME, Locatelli F, Fibbe WE. Mesenchymal Stromal Cells. Annals of the New York Academy of Sciences. 2009;1176(1):101–17. [DOI] [PubMed] [Google Scholar]
  • 20.Blancas-Mejia LM, Misra P, Dick CJ, Marin-Argany M, Redhage KR, Cooper SA, et al. Assays for Light Chain Amyloidosis Formation and Cytotoxicity Protein Misfolding Diseases: Springer; 2019. p. 123–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Dudakovic A, Camilleri E, Riester SM, Lewallen EA, Kvasha S, Chen X, et al. High-resolution molecular validation of self-renewal and spontaneous differentiation in clinical-grade adipose-tissue derived human mesenchymal stem cells. Journal of cellular biochemistry. 2014;115(10):1816–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Szklarczyk D, Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M, et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic acids research. 2016:gkw937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Metsalu T, Vilo J. ClustVis: a web tool for visualizing clustering of multivariate data using Principal Component Analysis and heatmap. Nucleic acids research. 2015;43(W1):W566–W70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Lin Y, Hogan WJ. Clinical Application of Mesenchymal Stem Cells in the Treatment and Prevention of Graft-versus-Host Disease. Advances in Hematology. 2011;2011:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Hafer-Macko CE, Dyck PJ, Koski CL. Complement activation in acquired and hereditary amyloid neuropathy. Journal of the Peripheral Nervous System. 2000;5(3):131–9. [DOI] [PubMed] [Google Scholar]
  • 26.Rodolico C, Mazzeo A, Toscano A, Pastura C, Maimone D, Musumeci O, et al. Amyloid myopathy presenting with rhabdomyolysis: evidence of complement activation. Neuromuscular Disorders. 2006;16(8):514–7. [DOI] [PubMed] [Google Scholar]
  • 27.Bodin K, Ellmerich S, Kahan MC, Tennent GA, Loesch A, Gilbertson JA, et al. Antibodies to human serum amyloid P component eliminate visceral amyloid deposits. Nature. 2010;468(7320):93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Matsuoka Y, Picciano M, Malester B, LaFrancois J, Zehr C, Daeschner JM, et al. Inflammatory responses to amyloidosis in a transgenic mouse model of Alzheimer’s disease. The American journal of pathology. 2001; 158(4):1345–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Shi Q, Chowdhury S, Ma R, Le KX, Hong S, Caldarone BJ, et al. Complement C3 deficiency protects against neurodegeneration in aged plaque-rich APP/PS1 mice. Science translational medicine. 2017;9(392):eaaf6295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Stoltzner SE, Grenfell TJ, Mori C, Wisniewski KE, Wisniewski TM, Selkoe DJ, et al. Temporal accrual of complement proteins in amyloid plaques in Down's syndrome with Alzheimer's disease. The American journal of pathology. 2000;156(2):489–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Eikelenboom P, Hack C, Kamphorst W, Rozemuller J. Distribution pattern and functional state of complement proteins and alpha 1-antichymotrypsin in cerebral beta/A4 deposits in Alzheimer's disease. Research in immunology. 1992;143(6):617–20. [DOI] [PubMed] [Google Scholar]
  • 32.Halle A, Hornung V, Petzold GC, Stewart CR, Monks BG, Reinheckel T, et al. The NALP3 inflammasome is involved in the innate immune response to amyloid-β. Nature immunology. 2008;9(8):857. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Masters SL, O’Neill LA. Disease-associated amyloid and misfolded protein aggregates activate the inflammasome. Trends in molecular medicine. 2011;17(5):276–82. [DOI] [PubMed] [Google Scholar]
  • 34.Diomede L, Romeo M, Rognoni P, Beeg M, Foray C, Ghibaudi E, et al. Cardiac light chain amyloidosis: the role of metal ions in oxidative stress and mitochondrial damage. Antioxidants & redox signaling. 2017;27(9):567–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Russo RC, Garcia CC, Teixeira MM, Amaral FA. The CXCL8/IL-8 chemokine family and its receptors in inflammatory diseases. Expert review of clinical immunology. 2014;10(5):593–619. [DOI] [PubMed] [Google Scholar]
  • 36.Kruger P, Saffarzadeh M, Weber AN, Rieber N, Radsak M, von Bernuth H, et al. Neutrophils: between host defence, immune modulation, and tissue injury. PLoS pathogens. 2015; 11(3):e1004651. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Kolaczkowska E, Kubes P. Neutrophil recruitment and function in health and inflammation. Nature Reviews Immunology. 2013;13(3):159. [DOI] [PubMed] [Google Scholar]
  • 38.Keeling J, Teng J, Herrera GA. AL-amyloidosis and light-chain deposition disease light chains induce divergent phenotypic transformations of human mesangial cells. Laboratory investigation. 2004;84(10):1322. [DOI] [PubMed] [Google Scholar]
  • 39.Teng J, Russell WJ, Gu X, Cardelli J, Jones ML, Herrera GA. Different types of glomerulopathic light chains interact with mesangial cells using a common receptor but exhibit different intracellular trafficking patterns. Laboratory investigation. 2004;84(4):440. [DOI] [PubMed] [Google Scholar]
  • 40.Horwitz EM, Andreef M, Frassoni F. Mesenchymal Stromal Cells. Current Opinion in Hematology. 2006;13(6):419–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Duijvestein M, Wildenberg ME, Welling MM, Hennink S, Molendijk I, van Zuylen VL, et al. Pretreatment with interferon-γ enhances the therapeutic activity of mesenchymal stromal cells in animal models of colitis. Stem cells. 2011;29(10):1549–58. [DOI] [PubMed] [Google Scholar]
  • 42.Pessina A, Bonomi A, Coccè V, Invernici G, Navone S, Cavicchini L, et al. Mesenchymal stromal cells primed with paclitaxel provide a new approach for cancer therapy. PLoS One. 2011;6(12):e28321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Bridge AJ, Pebernard S, Ducraux A, Nicoulaz A-L, Iggo R. Induction of an interferon response by RNAi vectors in mammalian cells. Nature genetics. 2003;34(3):263. [DOI] [PubMed] [Google Scholar]
  • 44.Salomon R, Staeheli P, Kochs G, Yen H-L, Franks J, Rehg JE, et al. Mx1 gene protects mice against the highly lethal human H5N1 influenza virus. Cell Cycle. 2007;6(19):2417–21. [DOI] [PubMed] [Google Scholar]
  • 45.Soscia SJ, Kirby JE, Washicosky KJ, Tucker SM, Ingelsson M, Hyman B, et al. The Alzheimer's disease-associated amyloid β-protein is an antimicrobial peptide. PloS one. 2010;5(3):e9505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Fulop T, Witkowski JM, Bourgade K, Khalil A, Zerif E, Larbi A, et al. Can an Infection Hypothesis Explain the Beta Amyloid Hypothesis of Alzheimer’s Disease? Frontiers in Aging Neuroscience. 2018;10(224). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplementary material
NIHMS1559004-supplement-supplementary_material.pdf (3.5MB, pdf)

RESOURCES